摘要
用较少的综合指标概括存在于大量观测数据中的各类信息,综合指标之间彼此不相关,各指标代表的信息不重叠的分析方法称为主成分分析。应用主成分分析法可实现对不同品牌啤酒风味差异性的评价、同一品牌啤酒风味一致性的评价、同一品牌不同生产厂之间一致性的评价和同一生产厂啤酒一致性的评价。该方法可消除各变量之间的共线性,减少变量的个数,结果直观;可以对啤酒样品风味的差异性、一致性与均一性进行分析比较。
The use of less composite indicators to generalize varieties of information in large amount of measurement data (no correlations among composite indicators and no information overlapping of each composite indicator) was called main components analysis. It could be used in evaluating the discrepancy of the flavor of different brands beer, the consistency of the flavor of the same brand beer, the consistency of the flavor of beer of the same brand but produced in different manufacturers, and the consistency of the flavor of beer produced in the same manufacturer. Main components analysis method could eliminate the colinear among all the variate and reduce variate quantity. It could be used in evaluating the discrepancy and the consistency of beer flavor.
出处
《酿酒科技》
北大核心
2007年第11期107-110,共4页
Liquor-Making Science & Technology
关键词
啤酒
主成分分析
啤酒风味
评价
beer
main components analysis
beer flavor
evaluation